Wikinvest Wire

Goldilocks math

Tuesday, February 12, 2008

Watching CNBC is not something that is done around here very often, but occasionally a few minutes of Power Lunch or Closing Bell will be tuned in just to see how things are going. Similarly, Andy Griffith reruns are favored over Kudlow, Cramer, and Fast Money since, even when skipping all the commercials, it's hard not to come away from these programs without feeling manipulated in one way or another.

Such was the case yesterday on Larry Kudlow's show when the discussion turned to the "near or already here" recession that the show's host has come to grudgingly accept as the reality on the ground.

"Recessions aren't such a bad thing", quipped Larry, "They've happened ten times in the postwar era and look what we have to show for it." A graphic was then shown (recreated below) and guests with opposing viewpoints were prodded to come around to the host's unwavering view that now is a good time to buy stocks.

Now, even with engineering math skills honed over years and years of school and work in private industry, yours truly still sometimes has a hard time with percentage gains, but that 86,677% number just looked freakishly large.

All the other figures seemed to make sense - over a 60 year period you get a lot of compounding, so a 3.4 percent GDP growth coming out to 634% over 60 years sounds reasonable.

In fact, to verify a proper understanding of how percentage gains are calculated you can go to the Wikipedia entry for Percentage and find out more than you ever wanted to know on the subject.

Here, the key passage as it applies to those big 60-year percent gains is:

Although percentages are usually used to express numbers between zero and one, any dimensionless proportionality can be expressed as a percentage. For instance, 111% is 1.11 and −0.35% is −0.0035.
So, the 634% increase in GDP would mean [insert] an incease of [/insert] 6.34 times the GDP in 1947 and the 87,667% gain in the S&P500 would mean [insert] an incease of [/insert] 876.67 times the stock market value at that same time.

That seems pretty simple, but those two numbers - 6.34 and 877 - seem to be way out of whack. Stocks went up at 138 times the rate of real economic growth?

It's easy enough to look up GDP, for example at Measuring Worth you'll find that real GDP in 1947 was $1574.5 billion, which, when multiplied by 6.34 yields $9,982 billion. Add the gain to the original value and you get somewhere close to the $10.1 trillion in the CNBC graphic.

But what about that S&P500 claim, the big number that Larry beat his guests over the head with in a never ending effort to encourage his American viewing audience to keep buying stocks?

What was the value of the S&P500 in 1947? Well, that's a hard question to answer because, according to this source, the S&P500 didn't exist until ten years later.
1957 -- The "416" becomes the Standard & Poor's 500 Composite Stock Price Index. Thanks to technological advancements, computers are introduced and permit the "500" to be calculated and disseminated at one-minute intervals throughout the trading day. In order to create a lengthy historical time series, the new "500" is linked to the 90 Stock Composite Price Index-daily S&P 500 Index prices become available back to 1928. The original "233" and "90" stock indices have evolved into the modern "500". The "500" now consists of 425 Industrials, 60 Utilities and 15 Rails, and has a base period of 1941-43=10.
Oh well, how about using the Dow Jones Industrial Average instead?

Finding the value of the Dow in 1947 is pretty easy - just go to Dow Jones and you can get these really cool charts, one for each decade.
So, it looks like it averaged about 180 during 1947, right around the time of the Marshall Plan and just prior to the invention of the tranisistor (these charts really are very good - I remember splicing them together and hanging them up on a wall sometime in the early 1990s).

So what would the Dow be today after a gain of 87,677% from 1947?

180 + 180 * 876.77 = 158,000

Oops! Now, that's real Goldilocks math!

The percent increase that they were probably going for was 8,767%.

What's an order of magnitude between friends?

AddThis Social Bookmark Button

10 comments:

Nick said...

No, their % is right... their methodology is suspect, but the # is correct. They are assuming constant 12% gains compounded over 60 years, which would be (1.12)^60, or 897.60, or 89,760% (the variance due presumably to a slightly different value than 12%).

It really just points to the inaccuracy of using 12% average annual gains as a measure of stock market performance, when averaging percentages doesn't work out right. For example, assume you had one year of -50%, and two years of +50%... you averaged (-50+50+50/3 = ) 16% over those 3 years. But in reality, you got (.5*1.5*1.5 = ) 12.5% return. The more extreme the gains and losses, the more wrong the average % is.

Hope that helps. :)

Anonymous said...

Don't forget to adjust for money supply growth. You'll see most of that yoy % gain is just the effect of more paper. Same goes for the other favorite, long-term real estate.

Show me a beast that comes by its living honestly and I will show you ten that intend to prey upon it.

Tim said...

Nick - you're probably right. They probably just said, "What's 12% a year for 60 years", then came up with 87,000%, which, unfortunately has nothing to do with the S&P500 performance.

Will said...

"So, the 634% increase in GDP would mean 6.34 times the GDP in 1947 and the..."

Wrong!

A 634% increase in anything would mean 7.34 times the original number.

A 100% increase is 2 times the original.
A 200% increase is 3 times the original.
...
...
A 634% increase is 7.34 times the original.

Tim said...

It's Right! now.
I told you I had problems with this.
A little slack please.

Anonymous said...

Calculated Risk had a similar sighting of CNBC math that rocketed past Kudlow's order-of-magnitude error into some parallel universe. Cramer, arguably the only host on CNBC more bombastic than Kudlow - was recently touting homebuilding stocks, claiming "homebuilders have basically stopped building -- they are building one-quarter of the homes they did in 2006 -- we are going to run out this year." Factually wrong on every point, according to CR.

Anonymous said...

Haven't you guys heard of Dividends? Reinvesting Dividends? There's your extra 5% per annum. In fact, over the long term, about 50% of your return in stocks comes from dividends and reinvesting them. So including dividends, you would expect to double the price-based return of the S&P or Dow. ahhh, amateurs.....

Anonymous said...

You should get your facts straight before you call someone an amateur:

Does the Dow Jones Industrial Average Quote include dividends reinvested?

Best Answer - It looks like lots of people answer questions that they don't understand. Perhaps they are confusing the DJIA with the S&P 500. That index does not include dividends.

Dividends are included in the DJIA. Whenever a dividend is paid, the divisor is adjusted. This is equivalent to reinvesting this in the index.

Source(s):
I offer myself as the source. Here are my qualifications.

1. PhD in Finance
2. Finance Professor
3. Former Dow Jones employee responsible for the system that collected stock quotations and computing the DJIA.

If you don't believe me, look at:

http://www.djindexes.com/mdsidx/index.cfm?event=showdow510Meth

Anonymous said...

Thanks JP, I already knew that.

I think you're trying to put too fine a point on this.

My point was that receiving dividends and reinvesting them boosts returns.

Anonymous said...

Didn't Goldilocks sign a DNR?

IMAGE

  © Blogger template Newspaper by Ourblogtemplates.com 2008

Back to TOP